DESCRIPTION: The applicant has generated a series of cellular models that mimic breast cancer's progression from estrogen (E2) dependence to E2-independence and to anti-estrogen cross resistance. Using 2-D gel analysis, the applicant has identified nucleophosmin (NPM) as a protein associated with acquired E2-independence. Data suggests that NPM levels are 8-fold higher in E2-independent and antiestrogen cross resistant cells when compared with E2-dependent cells. NPM can be detected in primary breast cancer specimens, and auto-antibodies to NPM can be detected in the sera of breast cancer patients. The applicant has three hypotheses: 1) elevated NPM expression and/or its state of phosphorylation are necessary and/or sufficient to confer the phenotypes associated with its expression; 2) the production of NPM autoantibody predicts recurrence or identifies tamoxifen non-responders; and 3) NPM levels in primary breast cancer specimens has prognostic significance. To test these hypotheses, three aims are proposed that will determine 1) the effects of overexpression and decrease expression of NPM on E2-independence and antiestrogen cross resistance, 2) NPM phosphorylation patterns in the various phenotypes, and 3) the predictive power of NPM and its autoantibodies using both retrospective and prospective studies of human clinical specimens.